基于改进遗传算法的传感器特性方程的建立

Establishment of the characteristic equation of the sensor using the genetic algorithm

在标准遗传算法的基础上,引入了变长度的编码方案、父子竞争的选择交叉机制、对优良个体的大变异操作等,算法的收敛速度得到很大提高,且能自动选择函数的形式、建立传感器的特性方程.结果表明,采用这种方法得到的拟合函数有着较高的精度.

The paper proposes a method for curve fitting based on genetic algorithm. It can choose the form of function to set up characteristic equation of sensor automatically in algorithm. The code scheme for changing length, the selection and crossover mechanisms of competition among the father and the son, greatly mutating for the fine individual etc. have been introduced on the basis of standard genetic algorithm. The convergence rate of the algorithm has been greatly enhanced. Simulation result shows that the fitting function obtained by this method has higher fitting precision.